Self-supervised Dynamic MRI Reconstruction

Mert Acar*, Tolga Çukur, İlkay Öksüz

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Citations (Scopus)

Abstract

Deep learning techniques have recently been adopted for accelerating dynamic MRI acquisitions. Yet, common frameworks for model training rely on availability of large sets of fully-sampled MRI data to construct a ground-truth for the network output. This heavy reliance is undesirable as it is challenging to collect such large datasets in many applications, and even impossible for high spatiotemporal-resolution protocols. In this paper, we introduce self-supervised training to deep neural architectures for dynamic reconstruction of cardiac MRI. We hypothesize that, in the absence of ground-truth data, elevating complexity in self-supervised models can instead constrain model performance due to the deficiencies in training data. To test this working hypothesis, we adopt self-supervised learning on recent state-of-the-art deep models for dynamic MRI, with varying degrees of model complexity. Comparison of supervised and self-supervised variants of deep reconstruction models reveals that compact models have a remarkable advantage in reliability against performance loss in self-supervised settings.

Original languageEnglish
Title of host publicationMachine Learning for Medical Image Reconstruction - 4th International Workshop, MLMIR 2021, Held in Conjunction with MICCAI 2021, Proceedings
EditorsNandinee Haq, Patricia Johnson, Andreas Maier, Tobias Würfl, Jaejun Yoo
PublisherSpringer Science and Business Media Deutschland GmbH
Pages35-44
Number of pages10
ISBN (Print)9783030885519
DOIs
Publication statusPublished - 2021
Event4th International Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2021 held in Conjunction with 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021 - Virtual, Online
Duration: 1 Oct 20211 Oct 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12964 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference4th International Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2021 held in Conjunction with 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
CityVirtual, Online
Period1/10/211/10/21

Bibliographical note

Publisher Copyright:
© 2021, Springer Nature Switzerland AG.

Funding

Acknowledgments. This paper has been produced benefiting from the 2232 International Fellowship for Outstanding Researchers Program of TUBITAK (Project No: 118C353). However, the entire responsibility of the publication/paper belongs to the owner of the paper. The financial support received from TUBITAK does not mean that the content of the publication is approved in a scientific sense by TUBITAK.

FundersFunder number
TUBITAK118C353

    Keywords

    • Cardiac MRI
    • Convolutional Neural Networks
    • Dynamic reconstruction
    • Self-supervised learning

    Fingerprint

    Dive into the research topics of 'Self-supervised Dynamic MRI Reconstruction'. Together they form a unique fingerprint.

    Cite this